fed-rag 0.0.19

Simplified fine-tuning of retrieval-augmented generation (RAG) systems.

FedRAG

---

FedRAG is an open-source framework for federated fine-tuning of Retrieval-Augmented Generation (RAG) systems.

Simplified RAG fine-tuning across centralized or federated architectures

Advanced RAG fine-tuning

Comprehensive support for state-of-the-art RAG fine-tuning methods that can be federated with ease.

Work with your tools

Seamlessly integrates with popular frameworks including HuggingFace, LlamaIndex, and LangChain — use the tools you already know.

Lightweight abstractions

Clean, intuitive abstractions that simplify RAG fine-tuning while maintaining full flexibility and control.

Installation

From package managers

# pypi
pip install fedrag

# conda-forge
conda install -c conda-forge fed-rag

[!NOTE] Extras for fed-rag are also available, such as the HuggingFace extra, which can be installed via pip install fed-rag[huggingface]

From source

git clone https://github.com/VectorInstitute/fed-rag.git
cd fed-rag

# install using pip
pip install -e .

# or, install using uv, our package manager tool of choice
uv sync --all-extras --group dev --group docs

Documentation

For more detailed documentation, visit our official documentation site.

• Getting Started
• Quick Starts
• Examples
• API Reference
• Glossary

[!TIP] This README provides a high-level overview, but our official documentation is updated more frequently with the latest features, tutorials, and API changes. For the most current information, please refer to the documentation site.

Examples

Check out our examples directory for more detailed usage examples:

• Basic RAG fine-tuning with federated learning
• Implementing RA-DIT with FedRAG
• Custom federated aggregation strategies
• Integration with popular LLM frameworks

Contributing

We welcome contributions! Please see our Contributing Guide for more details.

Citation

If you use FedRAG in your research, please cite our library:

@software{Fajardo_fed-rag_2024,
author = {Fajardo, Andrei and Emerson, David},
doi = {10.5281/zenodo.15092361},
license = {Apache-2.0},
month = mar,
title = {{fed-rag}},
url = {https://github.com/VectorInstitute/fed-rag},
version = {0.0.14},
year = {2025}
}

[!NOTE] The above citation may not reflect the most recent version of the library. We recommend using the Github citation widget (i.e. "Cite this respository") to obtain a citation entry reflecting the latest released version.

License

FedRAG is released under the Apache License 2.0.

Acknowledgements

FedRAG is developed and maintained by the Vector Institute.